First I believe that voice command for ATC and crew simulation is one of the most immersing factor along with a joke and throttle and TrackIR... I should add "when it works". In the following, let's suppose it works fine.In my opinion, voice recognition should be integrated within MS Flight and a part of the API should be dedicated to this. Basically, the API would let an addon register to the integrated speech recognition engine (SRE). Then, the engine would signal the registered addons whenever a speech has been processed. The addon is free to handle the signal in the usual way, using SimConnect (or its hopefully simplified successor). In this scenario the filtering of the relevant speech is done by the addon. The other and probably better way is to let the addon register the speechs that it can handle to the SRE. So far, so good.But what happends if two addons overlap in functionnalities and handle the same kind of speech (i.e. multicrew experience and fs2crew, an ATC addon for Asia and an other one for Europe) ? There is a need for a more clever way to dispatch the messages depending on the flight (aircraft, region of the world) so that two addons won't deal with the same speech and trigger conflicting events.There are severall benefits :- audio configuration done only once ;- voice learning (a painfull process) would be done only once ;- less work for voice user interfaces addon developpers ;- the engine could "understand" to some level the permutation of words in standard phraseology.What do you think ?

Voice recognition is already an integral part of Windows 7 and Vista. If flight was going to do voice recognition which I also think would be great it would probably involve that.

That's for sure. I did not say that Flight should implement a speech recognition engine but should *integrate* one. Microsoft Speech is probably a good choice and the one that every voice addon that I know use. My point was that there is a lot of code and user interface that could be "factorized".

The only problem with that is if there are improvements in the speech api itself, existing software will not be able to take advantage of it and you would have to wait until "aces" implements it. Not an ideal situation.

Assuming that third party devs are more responsive than the Flight team, that's right. Hopefully the update cycle for Flight will be probably much shorter (something like online update twice a year). A Game for Windows "Live" would be pretty dead with software updates every other year. But that's my wish, not a fact.

Many have been using voice since FS2004.

A voice api will totally improve the wider acceptability of flight. The ability to give voice commands to the 'co-pilot' like gear up and possibly a PTT (push to talk) feature for atc should be sellable by Microsoft.More importantly included in the SDK.

You can already do that with freeware and no SDKs since FS04.

Could you be more specific ? You mean by simulating mouse clicks ? Honestly ! The question is no what you can do or what you can't. The question is how to improve things. You can do plenty of funny things by hoocking FSX. The point here was to make voice recognition part of the product and gives it an easy, yet flexible, API.IMHO, the Flight team should focus on good APIs for all the subsystems (rendering, traffic, flight model, ATC...) so that they can be overriden by third parties. In my dreams, Flight is a framework (a standard structure that makes a flight simulator) and each piece have a default implementation (an improved version of the FSX one). In my deep dreams, EZCA would override the camera subsystem, Ultimate Traffic X would take over the traffic AI subsystem, VoxATC would provide a better ATC subsystem, Outerra will provide some fractal algorithms to the rendering engine, Austin will add some blade element theory flight model and so on...For 6 years, devs have been pushing the limits of the FSX SDK, some of the resulting addons were unexpected. It would be great if what are really "hacks" (or borderline use of the API) could be made a legitimate use and be seamlessy integrated in the Flight SDK (or should I say framework ?).

As mentioned, 'voice control' has been available in FS for a long time in various forms. The Speech API (SAPI) has been around since XP I think.FS2Crew offers voice control, along with other companies.I don't personally see Microsoft offering 'voice' as it would probably generate too many Support issues for them.Voice is great when it works, but it's complex, and there are numerous (user specific) things that can go wrong.Cheers,

Bryan, don't see any threat for your business here ! I'm only asking for an integration of a speech recognition engine (such as Speech) within the SDK, with a higher API level than the Speech one, so that every addon author that use voice recognition won't have to reinvent the wheel. This proposition is supposed to make your work easier !

'voice control' has been available in FS for a long time in various forms

Well, it's been available *along* FS for a long time. But it is time to get it *in* !If voice recognition is a bit fragile (you mention a lot of support), it's even more important to have it factorized and strengthen within Flight rather than have multiple buggy addons. Or is it the configuration that is problematic ? This could also be centralized and shared among addons (i.e. a control panel within Flight in the same vein that you have control panels for your HIDs).Did you try to run both FS2Crew and VoxATC (or any other voice interface addon) at the same time ? I suspect that it is problematic. But voice interface addons should coexist smoothly.As a devs, what are your wishes for a high level API for voice recognition within Flight SDK ? Refer to my first post for my ideas.

SamSam: I think you're missing the point when it comes to the use of Microsoft APIs. We need an SDK that interacts with the sim only, we can already incorporate every other API as we see fit. It would take the same amount of work either way. As BYork mentioned the problem isn't the code it's getting it to work flawlessly to meet users expectations. The only reason I or anyone else wouldn't rip off FSCrew is being we simply don't share the same passion for it, and he's got the market cornered already with multiple aircraft.

I may be wrong, but, to me, the voice command is on the same level than HIDs (Human Input Devices such as joystick, mouse...). Say you want to write an addon that computes some force feedback to your (dream) USB yoke. You would need the force applied on the airplane and the positions on the yoke axis. For the later, you may either (1) write a kernel driver for USB or (2) use the HID API of the Windows SDK, or simply (3) asks FS to give you the new positions of the HID (and hide all the complexity of enumerating, opening devices and having a thread that waits and reads the input devices). Which one do you prefer ?Again, how do you run multiple voice interface addons at the same time ? Who talked about riping off FSCrew ?

I may be wrong, but, to me, the voice command is on the same level than HIDs (Human Input Devices such as joystick, mouse...). Say you want to write an addon that computes some force feedback to your (dream) USB yoke. You would need the force applied on the airplane and the positions on the yoke axis. For the later, you may either (1) write a kernel driver for USB or (2) use the HID API of the Windows SDK, or simply (3) asks FS to give you the new positions of the HID (and hide all the complexity of enumerating, opening devices and having a thread that waits and reads the input devices). Which one do you prefer ?

Number 2... the code for 3 I've written so many times it wouldn't even take 30 minutes. On top of that I could use any existing .NET (30 or so) Lanagues if I stuck with number 2 :)

Bryan, don't see any threat for your business here ! I'm only asking for an integration of a speech recognition engine (such as Speech) within the SDK, with a higher API level than the Speech one, so that every addon author that use voice recognition won't have to reinvent the wheel. This proposition is supposed to make your work easier !Well, it's been available *along* FS for a long time. But it is time to get it *in* !If voice recognition is a bit fragile (you mention a lot of support), it's even more important to have it factorized and strengthen within Flight rather than have multiple buggy addons. Or is it the configuration that is problematic ? This could also be centralized and shared among addons (i.e. a control panel within Flight in the same vein that you have control panels for your HIDs).Did you try to run both FS2Crew and VoxATC (or any other voice interface addon) at the same time ? I suspect that it is problematic. But voice interface addons should coexist smoothly.As a devs, what are your wishes for a high level API for voice recognition within Flight SDK ? Refer to my first post for my ideas.

If Microsoft adds 'voice control' to Flight, it would probably be more of a good thing for FS2Crew than a threat because it will allow larger numbers of people to see just how good 'voice' can be.And FS2Crew's target demographic is sufficiently different from Microsoft's target demographic that they wouldn't be taking business from us. In general, FS2Crew targets the hard-core crowd; Microsoft targets the casual user.FYI, VoxATC and FS2Crew run side-by-side perfectly with no problems; I've done it personally. Cheers,